Instead of:

await page.locator("#main-content").scrollIntoView();

await page.click('button[data-testid="get-started-btn"]');

You write:

open playwright.dev

scroll all the way down

click on "Get started"

check if the page side menu is visible

.

This means:

- No more brittle selectors breaking on every refactor

- Non-technical stakeholders can write tests

- Tests survive UI refactors

- Much less maintenance overhead

.

It works with OpenAI, Claude (via compatible endpoints), or any local LLM that supports the OpenAI API format.

.

npm install e2e-test-agent

.

Would love to hear your thoughts! Especially curious about:

- What testing pain points this might solve for you

- Concerns about AI non-determinism in tests

- Ideas for extending it (API testing, database validation, etc.)

.

GitHub: https://github.com/armannaj/e2e-test-agent

npm: https://www.npmjs.com/package/e2e-test-agent

website: https://e2eagent.io